SUAVE: Unified Video-Action Models via Masked Diffusion

作者：Rhythm Syed, Jean Mercat, Sedrick Keh, Kushal Arora, Paarth Shah, Aykut Onol, Mengchao Zhang, Tony Dear · 单位：Columbia University, New York, NY, USA., Toyota Research Institute, Los Altos, CA, USA. · 会议/期刊：arXiv preprint · 方向：cs.RO · 发布日期：2026-10-06